Cytotoxic simplified tubulysin analogues

Article abstract of DOI:10.1021/jm701321p

An efficient route for the synthesis of the tubulysin family of antimitotic peptides was developed. Simplified tubulysin analogues were synthesized to define the minimum pharmacophore required for cytotoxicity. Simplified tubulysin analogues retain signif

Full text of DOI:10.1021/jm701321p

1530
J. Med. Chem. 2008, 51, 1530–1533
Letters
Chart 1a
Cytotoxic Simplified Tubulysin Analogues
Bhooma Raghavan,† Ranganathan Balasubramanian,†
Jaeson C. Steele,† Dan L. Sackett,‡ and Robert A. Fecik*,†
Department of Medicinal Chemistry, 8-101 WeaVer-Densford Hall,
308 HarVard Street S.E., UniVersity of Minnesota, Minneapolis,
Minnesota 55455-0353, and Laboratory of IntegratiVe and Medical
Biophysics, National Institute of Child Health and Human
DeVelopment, National Institutes of Health, 9 Memorial DriVe,
Building 9, Room 1E129, Bethesda, Maryland 20892-0924
a Mep, N-methylpipecolinic acid; Ile, isoleucine; Tuv, tubuvaline; Tup,
tubuphenylalanine.
ReceiVed October 19, 2007
Abstract: An efficient route for the synthesis of the tubulysin family
of antimitotic peptides was developed. Simplified tubulysin analogues
were synthesized to define the minimum pharmacophore required for
cytotoxicity. Simplified tubulysin analogues retain significant cytotox-
icity and reveal important preliminary structure–activity relationships.
Scheme 1. Synthesis of Tubuvaline Fragment 12
Tubulysins A and D (1 and 2, Chart 1) are representative
members of a family of antimitotic peptides isolated from the
myxobacteria Archangium gephyra and Angiococcus disciformis
by Höfle and co-workers.1,2 Tubulysin D retains potent anti-
cancer activity against the P-glycoprotein-expressing human KB-
V1 (multidrug-resistant cervix carcinoma) cell line (IC50 ) 0.31
nM),2 and tubulysin A is active against the HCT-15 (multidrug-
resistant colon carcinoma) cell line (GI50 ) 0.12 nM).3 Pre-
clinical studies with tubulysin A have also shown it to have
promising antiangiogenic effects.3 Tubulysin A noncompeti-
tively inhibits vinblastine binding to tubulin,4 suggesting that
the tubulysins bind to the peptide binding site located near the
Vinca alkaloid binding site of ꢀ-tubulin. The extraordinary
anticancer activity of the tubulysins against a validated target
makes them exciting leads for the development of novel drugs
for multidrug-resistant cancers.
Clinical development of the tubulysins has been hampered
by the extremely limited supply available from natural sources.
Fermentation of Angiococcus disciformis An d48 yields 0.25–1
mg/L tubulysin D,1,2 and fermentation of Archangium gephyra
Ar315 yields 3 mg/L tubulysin A.2 We sought to develop
efficient methods for the total synthesis of tubulysins that would
provide convenient access to synthetic analogues of potential
therapeutic interest and that would enable the study of their
interaction with tubulin. Various methods have been disclosed
for the total synthesis of tubulysin natural products,5–8 tubulysin
analogues,9–11 and key intermediates.12 We report here an
efficient total synthesis of a series of simplified tubulysin
analogues that were designed to establish the minimum structural
requirements for cytotoxicity. We hypothesized that the labile
acetyl and N,O-acetal functionalities of tubuvaline (Tuv) were
unnecessary for activity, and we also sought to examine the
effects of changes to the N-methylpipecolinic acid (Mep)
fragment.
1. Wolff rearrangement of diazomethylketone 313 under the
conditions of Limal and co-workers directly afforded Weinreb
amide 4.14 This was further protected as its bis-carbamate
derivative, Weinreb amide 5, for use as a potential acylating
agent. Our initial attempts at treating bromothiazole 615 with
organometallic reagents (including n-BuLi and i-PrMgCl) at low
temperature (-78 °C) followed by addition of Weinreb amide
5 led to complex mixtures resulting from addition of the
organometallic reagent to the ester of thiazole 6. Bromothiazole
6 was therefore reduced to alcohol 79 and protected as silyl
ether 8. We were gratified to find that treatment of silyl ether 8
with n-BuLi and Weinreb amide 5 smoothly afforded thiazolyl
ketone 9. Deprotection to alcohol 10, followed by a two-step
oxidation to aldehyde 11 and carboxylic acid 12 completed the
synthesis of a simplified tubuvaline fragment.
For the synthesis of tubuphenylalanine (Tup), we envisioned
that lactam 1316 could be diastereoselectively alkylated. Lactam
1317,18 and related lactams19 have previously been alkylated with
alkyl and benzyl bromides to give the trans products with modest
diastereoselectivity (∼4:1, trans/cis). Our initial attempts to
mimic these conditions with LDA and MeI gave modest yields
of monoalkylated lactam 14 with low diastereoselectivity (∼2:
1, trans/cis). Optimization of this alkylation using LiHMDS as
the base significantly improved the diastereoselectivity. Lactam
13 was treated successively with LiHMDS and MeI at -78 °C,
and the temperature and length of the reaction were varied
(Table 1). The major side product from these reactions was
Our synthetic route to tubuvaline, the central thiazole-
containing amino acid of the tubulysins, is depicted in Scheme
* To whom correspondence should be addressed. Phone: 612-626-6523.
Fax: 612-624-0139. E-mail: fecik001@umn.edu.
†
University of Minnesota.
National Institute of Child Health and Human Development.
‡
10.1021/jm701321p CCC: $40.75
2008 American Chemical Society
Published on Web 03/04/2008

Letters
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 6 1531
Scheme 2. Synthesis of Tubulysin Analogues
Table 1
entry
MeI (equiv)
time (h)
14 (%, dr)
15 (%)
1a
2
5
3
2
1.1
1.1
5
2
1.2
2
2
6, 10
15
6
4
1, 2
1, 2
45, 2:1
36, 2.3:1
35, 7:1
24, 12:1
14, 3:1
29, 3:1
30, 2:1
52, 10:1
NDb
26
27
30
27
27
28
12
3c
4
5d
6
7e
8c,f
a -78 to -20 °C. b ND: not determined. c -78 to -40 °C, warm to 23
°C. d 0.95 equiv of LiHMDS. e -78 to -40 °C, warm to -20 °C. f 1.2
equiv of NaHMDS (2.0M in THF).
dialkylated lactam 15, whose yield was not affected by changes
in the reaction temperature, time, or equivalents of LiHMDS
and MeI. A possible explanation for the significant yield of
dialkylated lactam 15 is that the intermediate enolate deproton-
ates the newly formed monoalkylated lactam 14 in competition
with alkylation by MeI. Under the optimal conditions for yield
and diastereoselectivity found here (entry 8) with NaHMDS,
the desired monoalkylated lactam 14 was obtained in 52% yield
and 10:1 dr, along with a 12% yield of dialkylated lactam 15.
The diastereoselectivity of the alkylations was determined by
1
analysis of the H NMR spectra of monoalkylated lactam 14.
These results compare favorably to the method of Wipf and
co-workers for installation of this stereogenic center by reduction
of an alkene with 3:1 diastereoselectivity.12
Table 2. Cytotoxicity of Tubulysin Analogs Against 1A9 Ovarian
Cancer Cells.a
Monoalkylated lactam 14 (as a 12:1 diastereomeric mixture)
was saponified to open the lactam and afford carboxylic acid
16, which was recrystallized to a single diastereomer (Scheme
2). The reduction of carboxylic acid 16 to alcohol 17 and
comparison of its spectral data to literature values12 confirmed
the stereochemistry of monoalkylated lactam 14 and carboxylic
acid 16. Alkylation of carboxylic acid 16 to give benzyl ester
18 followed by deprotection and coupling with the protected
carboxylic acid 12 yielded dipeptide 19. Removal of the N-Boc
groups and coupling with N-Boc-L-isoleucine gave the key
tripeptide 20. Alternative coupling conditions were explored but
these afforded tripeptide 20 in reduced yields.
To examine the effects of changes to the ring size, stereo-
chemistry, and methyl group of the N-methylpipecolinic acid
fragment on biological activity, tripeptide 20 was N-deprotected
and coupled to amino acids 21a–h (Scheme 2). Amino acids
21a–h that were not commercially available were synthesized
by published methods.5,20 Under the coupling conditions
required for N-methylamino acids 21a–d, the resulting tetrapep-
tides were contaminated with small amounts of dicyclohexylurea
(DCU). The partially purified products were subsequently treated
with LiOH to afford tetrapeptides 22a–d. N-Boc amino acids
21e–h were coupled with tripeptide 20 under different conditions
to give intermediate tetrapeptides 22e–h. Cleavage of the esters
and N-Boc groups afforded tetrapeptides 23a–d. This route
provided a series of simplified tubulysin analogues that lack
the N,O-acetal and have a ketone in place of the acetate found
in the natural products.
compd
R
n
abs config. (*) 1A9 IC50 (µM)
22a (FT-023)
22b (FT-025)
22c (FT-022)
22d (FT-024)
23a (FT-018)
23b (FT-020)
23c (FT-019)
23d (FT-021)
podophyllotoxin
HTI-286 (SPA110)
Me ·HCl
Me ·HCl
Me ·HCl
Me ·HCl
H ·HCl
1
1
2
2
1
1
2
2
R
S
0.8
18
R
S
0.2
25
R
S
R
S
>33
>33
23
H ·HCl
H ·HCl
H ·HCl
24
0.003
0.03 nM
a Data are the average of two independent IC50 value determinations
with a maximum drug concentration of 33 µM in DMSO.
cells (Table 2). Two tubulin polymerization inhibitors were used
as positive controls: the hemiasterlin analogue SPA110/HTI-
286 and podophyllotoxin. Several important structure–activity
relationships for the N-methylpipecolinic acid and tubuvaline
positions were revealed.
Comparison of 22a–d to 23a–d clearly shows a requirement
for a N-methyl group or tertiary amine at the N-terminus. There
is also a stereochemical requirement for the R-configuration (D-
amino acid configuration found in the natural products) at the
N-methylpipecolinic acid position. A potency difference of 2
orders of magnitude is observed between analogues that
incorporate a D-amino acid at this position (22a and 22c) versus
The initial cytotoxicity of tubulysin analogues 22a–d and
23a–d as IC50 values was measured against 1A9 ovarian cancer

1532 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 6
Letters
at the N-methylpipecolinic acid and tubuvaline positions. The
modifications to tubuvaline are notable, since they further
simplify its structure, and the N-methylpipecolinic acid changes
demonstrate the effects of stereochemistry and the N-methyl
group on activity. Two analogues, 22a (FT-023) and 22c (FT-
022), retain a potent level of cytotoxicity, and their antitubulin
activity is equipotent to HTI-286 and combretastatin A4.
Additional modifications to the tubulysin scaffold and further
biological properties of analogues 22a and 22c will reported in
due course.
Acknowledgment. This research was generously supported
by Grant RSG-05-105-01-CDD (to R.A.F.) from the American
Cancer Society and was supported in part by intramural funds
from the National Institute of Child Health and Human
Development, NIH. The authors thank Beverly G. Ostrowski
for assistance with the NMR instruments. Funding for NMR
instrumentation in the Department of Biochemistry, Molecular
Biology, and Biophysics was provided by the University of
Minnesota Medical School, NSF (Grant BIR-961477), and the
Minnesota Medical Foundation.
Figure 1. Inhibition of tubulin polymerization by 22a and 22c. Rat
brain tubulin (10 µM) was mixed with various concentrations (0.5, 1,
2, and 4 µM) of test compounds on ice in 0.8 M sodium glutamate,
pH 6.6, and 0.4 mM GTP. Samples were warmed to 33 °C, and optical
density at 350 nm was followed with time in a SpectraMax Plus plate
reading spectrophotometer (Molecular Devices, Sunnyvale, CA). Only
one concentration is shown for clarity, 1 µM for all compounds: [ )
control, 2 ) 22a, 9 ) 22c, × ) HTI-286. Combretastatin A4 is not
shown.
Supporting Information Available: Experimental procedures
and spectral data for 4, 5, 7–12, 14–16, 18–20, 22a–h, and 23a–d;
1H and 13C NMR spectra for 4, 5, 8–12, 14–16, 18–20, 22a–h,
and 23a–d; and experimental procedures for biological testing. This
material is available free of charge via the Internet at http://
pubs.acs.org.
those with an L-amino acid (22b and 22d). The ring size of
N-methylpipecolinic acid can be varied, provided the N-methyl
group and stereochemical requirements are followed. Com-
pounds incorporating N-methyl-D-proline (22a) and N-Me-D-
pipecolinic acid (22c) are roughly equipotent.
At the tubuvaline position, significant modifications are
tolerated. The cytotoxicity of 22a and 22c shows that a tertiary
amide is not required, and the acetoxy or hydroxy groups present
in the natural products can be replaced with a ketone. The latter
change is significant, since it eliminates the need for a
stereogenic center that is challenging to install.
The most active compounds in this series, 22a (FT-023) and
22c (FT-022), have potent cytotoxicity against 1A9 ovarian
cancer cells; however, they are notably less cytotoxic than HTI-
286, a hemiasterlin analogue presently in phase I clinical trials.
To confirm their mechanism of action, analogues 22a and
22c were also tested for the ability to inhibit tubulin polymer-
ization in vitro (Figure 1).21 Under these conditions, the IC50
obtained was approximately 1.0 µM for 22a, 1.8 µM for 22c,
1.0 µM for HTI-286, and 1.1 µM for combretastatin A4. These
results establish inhibition of tubulin polymerization as the
mechanism of tubulysin analogues 22a and 22c.
Interestingly, the relatively equivalent IC50 values of 22a and
22c to HTI-286 in the tubulin polymerization assay do not
correlate to their cytotoxicity. The reasons for this discrepancy
are not known at this time. One potential explanation is that
22a and 22c have decreased permeability across the cell
membrane.
These results confirm and extend the preliminary structure–
activity data reported for two small series of tubulysin D11 and
H10 analogues. The cytotoxicity and tubulin inhibition data for
22a and 22c are comparable to those found for two other
tubulysin analogues, although different cell lines and assay
conditions were used.10 We show here that additional modifica-
tions to tubuvaline are tolerated. These changes, the lack of a
tertiary amide, and elimination of a stereogenic center simplify
the processes required for the synthesis of tubulysin analogues.
Synthesis of 22a and 22c was achieved in 12 linear steps from
commercially available materials, which compares favorably to
the total syntheses of natural tubulysins.22
References
(1) Sasse, F.; Steinmetz, H.; Heil, J.; Höfle, G.; Reichenbach, H.
Tubulysins, new cytotoxic peptides from myxobacteria acting on
microtubuli. Production, isolation, physico-chemical and biological
properties. J. Antibiot. 2000, 53, 879–885.
(2) Steinmetz, H.; Glaser, N.; Herdtweck, E.; Sasse, F.; Reichenbach, H.;
Höfle, G. Isolation, crystal and solution structure determination, and
biosynthesis of tubulysinsspowerful inhibitors of tubulin polymeri-
zation from myxobacteria. Angew. Chem., Int. Ed. 2004, 43, 4888–
4892.
(3) Kaur, G.; Hollingshead, M.; Holbeck, S.; Schauer-Vukasinovic, V.;
Camalier, R. F.; Dömling, A.; Agarwal, S. Biological evaluation of
tubulysin A: a potential anticancer and antiangiogenic natural product.
Biochem. J. 2006, 396, 235–242.
(4) Khalil, M. W.; Sasse, F.; Luensdorf, H.; Elnakady, Y. A.; Reichenbach,
H. Mechanism of action of tubulysin, an antimitotic peptide from
myxobacteria. ChemBioChem 2006, 7, 678–683.
(5) Peltier, H. M.; McMahon, J. P.; Patterson, A. W.; Ellman, J. A. The
total synthesis of tubulysin D. J. Am. Chem. Soc. 2006, 128, 16018–
16019.
(6) Dömling, A.; Beck, B.; Eichelberger, U.; Sakamuri, S.; Menon, S.;
Chen, Q.-Z.; Lu, Y.; Wessjohann, L. A. Total synthesis of tubulysin
U and V. Angew. Chem., Int. Ed. 2006, 45, 7235–7239.
(7) Dömling, A.; Beck, B.; Eichelberger, U.; Sakamuri, S.; Menon, S.;
Chen, Q.-Z.; Lu, Y.; Wessjohann, L. A. Total synthesis of tubulysin
U and V. Angew. Chem., Int. Ed. 2007, 46, 2347-2348 (Corrigendum).
(8) Sani, M.; Fossati, G.; Huguenot, F.; Zanda, M. Total synthesis of
tubulysins U and V. Angew. Chem., Int. Ed. 2007, 46, 3526–3529.
(9) Wipf, P.; Wang, Z. Total synthesis of N14-desacetoxytubulysin H. Org.
Lett. 2007, 9, 1605–1607.
(10) Wang, Z.; McPherson, P. A.; Raccor, B. S.; Balachandran, R.; Zhu,
G.; Day, B. W.; Vogt, A.; Wipf, P. Structure-activity and high-content
imaging analyses of novel tubulysins. Chem. Biol. Drug Des. 2007,
70, 75–86.
(11) Patterson, A. W.; Peltier, H. M.; Sasse, F.; Ellman, J. A. Design,
synthesis, and biological properties of highly potent tubulysin D
analogues. Chem.sEur. J. 2007, 13, 9534–9541.
(12) Wipf, P.; Takada, T.; Rishel, M. J. Synthesis of the tubuvaline-
tubuphenylalanine (Tuv-Tup) fragment of tubulysin. Org. Lett. 2004,
6, 4057–4060.
(13) Guibourdenche, C.; Podlech, J.; Seebach, D. Nucleosides, with
intermediates generated by Wolff rearrangement of amino acid derived
diazo ketones: Preparation of ꢀ-amino acid derivatives. Liebigs Ann.
1996, 1121–1129.
In summary, we report here the synthesis and initial biological
evaluation of eight tubulysin analogues that incorporate changes
(14) Limal, D.; Quesnel, A.; Briand, J.-P. Direct synthesis of N-protected
ꢀ-amino dimethylhydroxamates: application to the solid-phase

Letters
synthesis of a peptide incorporating a new amide bond surrogate
ψ[CH2CH2NH]. Tetrahedron Lett. 1998, 39, 4239–4242.
(15) Kelly, T. R.; Lang, F. Total synthesis of dimethyl sulfomycinamate.
J. Org. Chem. 1996, 61, 4623–4633.
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 6 1533
(19) Ezquerra, J.; Pedregal, C.; Rubio, A.; Vaquero, J. J.; Matía, M. P.;
Martín, J.; Diaz, A.; Navío, J. L. G.; Deeter, J. B. Stereoselective
double alkylation of ethyl N-Boc-pyroglutamate. J. Org. Chem. 1994,
59, 4327–4331.
(16) Smrcina, M.; Majer, P.; Majerová, E.; Guerassina, T. A.; Eissenstat,
M. A. Facile stereoselective synthesis of γ-substituted γ-amino acids
from the corresponding R-amino acids. Tetrahedron 1997, 53, 12867–
12874.
(17) Salituro, F. G.; Baker, C. T.; Court, J. J.; Deininger, D. D.; Kim, E. E.;
Li, B.; Novak, P. M.; Rao, B. G.; Pazhanisamy, S.; Porter, M. D.;
Schairer, W. C.; Tung, R. D. Design and synthesis of novel
conformationally restricted HIV protease inhibitors. Bioorg. Med.
Chem. Lett. 1998, 8, 3637–3642.
(18) Kazmierski, W. M.; Andrews, W.; Furfine, E.; Spaltenstein, A.; Wright,
L. Discovery of potent pyrrolidine-based HIV-1 protease inhibitors
with enhanced drug-like properties. Bioorg. Med. Chem. Lett. 2004,
14, 5689–5692.
(20) Khalil, E. M.; Subasinghe, N. L.; Johnson, R. L. An efficient and high
yield method for the N-tert-butoxycarbonyl protection of sterically
hindered amino acids. Tetrahedron Lett. 1996, 37, 3441–3444.
(21) Hamel, E. Evaluation of antimitotic agents by quantitative comparisons
of their effects on the polymerization of purified tubulin. Cell Biochem.
Biophys. 2003, 38, 1–21.
(22) The total synthesis of tubulysin D was completed in 16 linear steps
(ref 5). Tubulysins U and V required 15 and 16 linear steps,
respectively (ref 8). These natural products were also synthesized in
10 linear steps; however, the penultimate intermediate completely
epimerized (ref 6).
JM701321P